Research On Digital Track Map Aided BeiDou/INS Deeply Integrated Train Positioning Method

It’s crucial to obtain the train location accurately and timely for train control system and the accurate train location will guarantee the safety of train operation. With the rapid development of BeiDou navigation satellite system (BDS) in China, the current trends aim at integrated train positioning based on BDS in order to improve rail transportation efficiency and safety performance. As more sensors are adopted in integrated train positioning system, a more complicated architecture, which can more deeply fuse the different sources of measurement information to realize a tight integration, should be designed to obtain a more high accuracy.In this thesis, according to the technical characteristics of BDS and integrated train positioning, to solve some key problems concerning multi-sensor information fusion for safe train positioning, acquisition and adaptive tracking algorithm for BeiDou B1/B2 civil signals are studied. Based on multi-sensors information fusion, a novel digital track map (DTM) aided BeiDou/INS deep integration strategy is proposed, and a new nonlinear filtering algorithm called the cubature information filter (CIF) for deep integration is presented. Furthermore, fault detection architecture is constructed for deeply integrated train positioning system.The innovations of the thesis are as follows:(1) A fine acquisition algorithm based on short time matched filter (STMF) and fast Fourier transform (FFT) with simplified differentially coherent integration (SDCI) for BeiDou B1/B2 civil signals is proposed. This algorithm improves the detection Signal-to-Noise Ratio (SNR) of BeiDou signals when the integration time is limited. Meanwhile, the algorithm is also developed under hardware platform easily and has higher accuracy of acquisition.(2) An adaptive tracking loop with double-filter model based on innovation covariance for BeiDou B1/B2 civil signals to enhance the adaptive ability of tracking loop is proposed. The proposed algorithm is adaptable to signal strength change, and then adjusts equivalent noise bandwidth to track signals stably. This algorithm solves the problem that receiver can easily lose lock in harsh environments along railway (signals attenuation or refraction).(3) This thesis presents a novel DTM aided BeiDou/INS correction model with dual-feedback tracking loop which is based on BeiDou/INS deep integration model with double-filter for train positioning. Within the tracking loop, an adaptive tracking loop with double-filter model is used for the code and carrier tracking closed-loop. By means of fusing DTM measurement equation into integrated filter which reduce the impact of inaccuracy of INS, the correction of code tracking error and carrier frequency tracking error are finished among tracking channels. The channel filter and integrated filter estimate and correct the errors respectively, which enhance positive correction among tracking channels, and obtain higher tracking accuracy and train positioning accuracy. Meanwhile, this thesis presents an augmented state model based on DTM aided BeiDou/INS deeply integrated train positioning system to correct map measuring errors during the process of filtering.(4) A new nonlinear filtering algorithm called the CIF is presented for deep integration system. The proposed algorithm has higher nonlinear estimation accuracy. Meanwhile, because the filter algorithm is easy to extend for multi-sensor state estimation without more computational complexity, and then fault detection architecture is designed for deeply integrated train positioning system.To evaluate the performance of deeply integrated train positioning system in this thesis, the DTM aided BeiDou/INS deeply integrated train positioning software is designed and a simulation platform is also constructed. The sensor error models, train trajectory, satellite signal power adjustment and satellite faults setting are realized to validate theoretical model. Meanwhile, Qinghai-Tibet railway field experiments are finished to further test theoretical model. Test results demonstrate that the proposed DTM aided BeiDou/INS deeply integrated positioning method can improve train positioning accuracy. Theoretical results in this thesis will lay the foundation for BDS-based train control system.